37 results about "Movie reviews" patented technology

The invention discloses a text classification method combining dynamic word embedding with part-of-speech tagging, and provides the text classification method based on a deep neural network through combining dynamic word embedding with part-of-speech tagging. The method can fully utilize the advantages that a large-scale corpus can provide more accurate grammar and semantic information, and can also adjust word embedding by combining with the features of the corpus during a model training process, and thus the features of the corpus can be better learned. Meanwhile, classification accuracy can be further improved by combining with part of speech information of words in sentences. The invention also comprehensively utilizes the advantages of LSTM in the aspect of learning context information of words and part of speech in the sentences, and the advantages of CNN in the aspect of learning text local features. The classification model provided by the invention has the advantages of high accuracy and strong universality, and achieves good effect in some famous public corpuses including IMDB corpus, Movie Review and TREC.

Systems and methods are provided for communicating filtered event program information and broadcasting user searchable event program information to vehicles. For example, there is provided a system for broadcasting event program information to at least one vehicle that is filtered according to the vehicle's current location. In one embodiment, the system comprises one or more antennas, a broadcast receiver unit having a data filter, and a navigational device that displays event program information by venue within a set distance from the vehicle's current location. The display further comprises multi-leveled sub-menus that provide a user with a variety of movie related information such as screening times, cast listings, movie reviews, movie trailers, etc. The system further comprises a memory buffer unit in the receiver unit that stores the vehicle's last reported location data and cross-referenced databases of movie program information periodically updated so as to minimize bandwidth requirements of the system.

The embodiment of the invention provides a short English movie review classification method and device, and the method comprises the steps: obtaining a feature word set of a to-be-classified short English movie review text; performing prediction classification on the feature word set through a pre-trained first-layer classification model to obtain a first prediction result set, splicing the firstprediction result set, and taking arithmetic mean to obtain a new test set; predicting the new test set through a pre-trained second-layer classification model to obtain a second prediction result set, and minimizing the second prediction result set; performing maximum normalization processing to obtain a normalized result set; and processing the normalized result set by adopting a simple majoritymethod to obtain a classification result of the to-be-classified short English movie review. According to the technical scheme of the invention, the accuracy and robustness of the classification result are greatly improved.

The invention provides a neural collaborative filtering model recommendation method based on lambda Mart, and the method comprises the following steps: S1, inputting user information which comprises user basic information and movie comment information, and the movie comment information comprises scored movie information and unscored movie information; S2, enabling the embedding layer to map the user information into a user feature vector, and mapping the movie comment information into a movie feature vector; S3, inputting the user feature vector and the movie feature vector into a neural collaborative filtering model, extracting high-order feature information, and extracting sorting information at the same time; and S4, processing the high-order feature information and the sorting information to obtain a recommendation result, and outputting the recommendation result. According to the method, an LMNCF model is put forward, a neural collaborative filtering model is improved, implicit high-order feature information is extracted through nonlinear feature processing of a multi-layer perceptron, sorting information is extracted through a lambda Mart personalized sorting algorithm, and recommendation is more accurate.

The invention discloses a movie comment sentiment analysis method based on document vectors. The method comprises the following steps: firstly, carrying out data cleaning and preprocessing on movie comments to construct an emotion analysis data set; and training an improved document vector model by using comments of the movie and scores corresponding to the comments; inputting the movie comments into an improved document vector model to obtain word vectors and document vectors of the movie comments, averaging the word vectors corresponding to the movie comments, and splicing the word vectors with the document vectors to generate feature vectors; and finally, performing emotion classification on the movie comments by using the feature vectors generated based on the movie comments and the score training classification model corresponding to the movie comments. By using the improved document vector generation method, the movie comment sentiment classification accuracy is improved.

The invention relates to a movie recommendation system for relieving data sparsity. The movie recommendation system comprises an initial data input module and a recommendation list module, the initialdata input module comprises movie introduction data, movie comment data, actual scoring data and theater group personnel information data; the movie introduction data and the actual scoring data areprocessed by an improved convolutional neural network and then transmitted to a recommendation list module; establishing a mathematical model according to the theater group personnel information data,and embedding the mathematical model into the recommendation list module; and the movie comment data is processed through an algorithm and then transmitted to a recommendation list module. The invention provides a movie recommendation system and method for relieving data sparsity. The movie introduction data, the movie comment data, the actual scoring data and the theater group personnel information data are collected, the improved convolutional neural network and the mathematical model are utilized to perform movie information fusion, the new recommendation list is generated, the accuracy ofthe system is improved, and the problem of data sparsity of the user-scoring matrix is relieved.

The invention provides a text classification method of a gating loop unit based on residual jump connection. The method is characterized in that on the basis of an existing gating loop unit, residual information is introduced by utilizing jump connection and gating limitation of an expressway network is combined, and meanwhile, an unsaturated activation function and batch standardization are used to improve a definition formula of the gating loop unit, so that the gating loop unit can capture an ultra-long-term dependency relationship while coping with the gradient disappearance problem, and residual information is limited not to occupy a main position by utilizing gating of the expressway network. According to the method, the trained network model is utilized to automatically classify the Chinese texts and classify the positive/negative evaluation of the movie review, so that the labor cost is reduced. A plurality of groups of contrast experiments prove that the method has a certain applicability in a text classification task.

The invention provides an algorithm for automatically generating movie labels based on movie reviews. The algorithm in the invention fully considers the missing problem existing in the data set of thecurrently labeled movies. Firstly, an unsupervised algorithm with weights is adopted to automatically replenish the labels for the training set from the movie reviews. At the same time, the inventionalso fully considers the relationship between the similarity of each attribute and the similarity of labels of two movies, and predicts the mapping from each attribute to the similarity of labels bya machine learning method, instead of adopting simple similarity such as cosine similarity to calculate the rough similarity relationship. Finally, after the candidate multisets of labels are obtainedby using the traditional K-nearest neighbor algorithm, the method does not use simple evaluation criteria to sort and select the label set, but uses the graph algorithm based on label co-occurrence relationship to determine the order of the candidate labels, thus deciding the final label set.

The invention provides a personalized movie recommendation method and device and electronic equipment, and the method comprises the steps of determining a movie label set of a plurality of movies in advance, wherein the movie label set comprises a plurality of movie labels; generating a user label set of the user according to the viewing record of the user, wherein the user label set comprises a plurality of user labels; determining the overall association degree between the user and the movie according to the association degree between the plurality of user tags and the plurality of movie tags; and selecting one or more movies from the plurality of movies as target movies according to the overall association degree of the plurality of movies, and pushing the target movies to the user. According to the personalized movie recommendation method and device and the electronic equipment provided by the embodiment of the invention, the personalized movie label is formed on the basis of the movie review information, the overall matching degree between the user and each movie can be determined more finely and accurately, accurate recommendation can be realized in a targeted manner on the basis of the characteristics of the user, and personalized movie recommendation is realized.

The invention discloses a method for pushing a movie review based on video playing time and a digital television. The invention belongs to the field of internet, and provides a method for pushing a movie review based on video playing time and a digital television. The method includes the following steps: recording the matching relation between the video playing time and a movie review segment; when a user watches a movie, detecting the video playing time; if the current time is the video playing time which matches the movie review segment, pushing the movie review to the user. According to the invention, the method herein, by recording the matching relation between the video playing time and the movie review segment, enables the user to browse movie reviews that match the plot of the movie when the user watches the movie, deepens user's understanding of the movie and improves user experience in terms of watching movie.

The invention discloses a movie review information retrieval system and method based on emotion analysis. The system comprises: a comment information search module, a comment information crawling module, a comment information cleaning module, a comment information extraction module, a comment information analysis module, a comment information recommendation module and a comment information viewingmodule. By crawling the movie-related comment information on the website and cleaning and extracting the comment information, the emotion and quality of comments can be analyzed, so as to enable theuser to accurately select the high-quality movie and improve the quality of the user view through viewing the movie comments.

The invention discloses a data mining and processing method and system based on multimedia digital image review, which comprises: using web crawler technology to mine public film review data for at least one movie on social networks and carrying out a second mining of a first preset amount on user page data for publishing the film review data while mining the film review date; establishing the mapping relationship between the movie review data and the user page data; carrying out analysis of interest attributes for the user page data and obtaining interest attribute points in turn, classifying and aggregating the user page data whose interest attribute points matching quantity exceeds a second preset amount to form several user page data sets; detecting pending operation events, and mining and executing data which is opened by users of the pending operation events on the social networks, carrying out analysis of interest attributes for the data and obtaining interest attribute points in turn, and finding the user page data set whose the interest attribute point matching quantity exceeds the second preset amount. The data mining and processing method and system based on multimedia digital image review enable users to find valuable information more conveniently and quickly.